1 |
On 2018-04-28 12:00 AM, Michał Górny wrote: |
2 |
> W dniu pią, 27.04.2018 o godzinie 17∶56 -0700, użytkownik Patrick McLean |
3 |
> napisał: |
4 |
>> On 2018-04-27 09:05 AM, Mike Gilbert wrote: |
5 |
>>> On Fri, Apr 27, 2018 at 9:03 AM, Michał Górny <mgorny@g.o> wrote: |
6 |
>>>> |
7 |
>>>> Firstly, I'd like to do two changes simultaneously to reduce --newuse |
8 |
>>>> rebuilds: |
9 |
>>>> |
10 |
>>>> a. switching from CPython 3.5 to 3.6, |
11 |
>>>> |
12 |
>>>> b. disabling CPython 3.4. |
13 |
>> |
14 |
>> I assume you mean remove from the default PYTHON_TARGETS, not disable it |
15 |
>> completely? I would wait until after python 3.4 is no longer getting |
16 |
>> updates before removing it from the eclass and the tree. |
17 |
> |
18 |
> I meant disabling it completely. The default is 3.5 for a long time |
19 |
> already, and there is no reason to expect people to go back and test |
20 |
> their packages against 3.4. The plan was to support only one Python 3.x |
21 |
> release behind current stable.>>>> |
22 |
>>>> I'm thinking of a soft deadline on 2018-06-01, i.e. giving developers |
23 |
>>>> a full month to prepare. If things don't go well, we can always |
24 |
>>>> postpone it. |
25 |
>>>> |
26 |
|
27 |
IMHO, it would be better to wait until upstream deprecates Python 3.4 |
28 |
before we disable/drop it. There are downstreams (my employer for one, |
29 |
and I am sure we are not the only ones) that are still using it, and |
30 |
have plans in place to migrate by the upstream deprecation date. |
31 |
|
32 |
>>>> According to my lists, we only have 6 packages relying on py3.4 right |
33 |
>>>> now [1] and no pending stabilizations for that. I will report bugs for |
34 |
>>>> those packages today. |
35 |
>>>> |
36 |
>>>> The list for 3.5->3.6 migration is longer [2]. However, it seems that |
37 |
>>>> many of those packages are rather isolated [3] and apparently |
38 |
>>>> unmaintained. Apparently the biggest targets are OpenStack |
39 |
>>>> and Flask. It's all doable. |
40 |
>>>> |